Finding vaccine-induced immune response biomarkers that predict vaccine-protection against HIV-1 and other signi?cant pathogen infection, and understanding their role as mechanisms of protection, is an essential part of vaccine research. We distinguish three stages in immune correlates studies. The ?rst stage deals with measurement of immune response and may involve experimental design. The second stage performs preprocessing of immune response variables obtained in the ?rst stage to form composite immune response variables. The third stage examines the association between clinical outcome and immune response variables. We will develop novel statistical methods for each stage in order to increase the signal- to-noise ratio in immune response measurement, create immune response scores with maximum signal diversity, and increase the power of hypothesis testing by using change point regression model.